1 Reply Latest reply on Nov 21, 2019 4:00 AM by Srinath Menon-Oracle

    IBR Conversions slow in large cluster

    Whoopass

      We have a client who has 3 WCC 11g nodes (running on Linux) on one cluster and multiple Windows based clusters with IBRs, as follows:

       

      1 Cluster with 24 IBR nodes

      2nd Cluster with 8 IBR nodes

      3rd Cluster with 8 IBR nodes

       

      Some of the nodes at the end of the first cluster are converting very slowly. Could having so many (24) IBR nodes on one cluster affect the performance of the IBR nodes on this cluster? Would breaking off some of those nodes onto the separate cluster help performance?

       

      Any other ways to improve the performance of Windows based IBR servers?

       

      Has anybody run into performance issues with a similar configuration, as above?

       

      Let me know, please.

       

      Thanx,

       

      Chad :^)

        • 1. Re: IBR Conversions slow in large cluster
          Srinath Menon-Oracle
          Some of the nodes at the end of the first cluster are converting very slowly. Could having so many (24) IBR nodes on one cluster affect the performance of the IBR nodes on this cluster? Would breaking off some of those nodes onto the separate cluster help performance?

          Conversions could be slow due to various reasons like conversion engine performance , file sizes and complexities etc. But, to find what is causing the issue you will need to enable tracing requestaudit,file*,ref*,ibr* + Full verbose.

          Along with this Enable verbose logging for the agent from IBR Console - Administration - Agent Management .

           

          Plus, look for latest OIT patches that are available for the version and platform and apply the same on this environment to test further .

          1 person found this helpful